I’ve seen some coverage of this new putative Jurassic flower published online and open-source in Historical Biology a few weeks ago, so I decided I’d share my thoughts. If you didn’t hear about it, here are some links to the open source article and some associated media:
First, some background. The oldest angiosperms that can be assigned to living groups are Cretaceous. These are plants that, if you saw them alive, anyone who has taken a botany class would recognize immediately as angiosperms. If we assume that the fossil record as we know it tells a reliable story, then the story of the angiosperm diversification took place during the Cretaceous.
Diversification of Potomac Group angiosperms
However, flowering plants are related to other plants, and the other living seed-plant groups to which angiosperms are related (albeit distantly) are conifers, Ginkgo, cycads, and Gnetales. These groups have fossil records that go back much farther than the Cretaceous. This means that the origin of angiosperms is more uncertain. There are extinct seed-plants that lived during the Jurassic and Triassic that are more closely related to angiosperms than to any of the living gymnosperms. These extinct groups would have had some, but not all of those features that unite modern angiosperms and distinguish them from gymnosperms. Some of these extinct angiosperm relatives are certainly still in the rocks waiting to be discovered, but others are probably already in our collections in museums around the world. Most of us paleobotanists have our opinions and hypotheses about which known groups of extinct gymnosperms are angiosperm-relatives; but I wouldn’t say there is consensus yet and there is still so much potential to discover well-preserved highly informative fossils in the field.
As for this recent paper, authors Liu and Wang document what they interpret as an incomplete fossil flower that had 5 sepals, 5 petals, probably 5 stamens, and a gynoecium (pistil) with a unilocular ovary, unitegmic ovules, and a hairy style. Their interpretation of the characters placed the fossil among the living eudicot angiosperms rather than near the base of angiosperm phylogeny, but they did not include a phylogenetic analysis. They also argue that this fossil is mid-Jurassic and therefore one of two things should follow: 1) either modern angiosperms (including groups like monocots, magnoliids, and water lilies) are all many tens millions of years older than we think based on the rest of the fossil record, or 2) the consensus regarding angiosperm phylogeny that has been reached over the last ~20 years of molecular phylogenetics is wrong. While either of these are possible, they are unlikely because most of the evidence of angiosperm evolution fits a different model much better. As well-preserved fossils of modern angiosperm orders and families appear in the Cretaceous fossil record, they do so in a sequence that matches what we expect based on our knowledge of angiosperm relationships, and our knowledge of angiosperm relationships comes from DNA, not from fossils. Furthermore, the semi-independent fossil record of leaves, pollen, and charcoalified flowers all tell the same story of a Cretaceous diversification.
Summary of angiosperm phylogeny
As for their interpretation of the morphology of this plant fossil, I have to say I am just not convinced. I see the structures labeled as sepals and petals, but I think they look more likely they are helically arranged rather than whorled like a pentamerous eudicot flower, and while I see some parallel striations, and I don’t see any indication of venation. As for the tetrasporangiate anthers, I don’t see them. The argument appears to be based on the position of these small structures above the petals, and on the constriction in the middle of these small structures. I don’t see the different layers of the microsporangia and the detail view of the pollen grain inside one of the “anthers” doesn’t look like pollen to me. The pollen of eudicot angiosperms is generally both tough and very distinctive. Finally, as for the gynoecium, while I see the thin structure they interpret as a hairy style, I am not convinced by the illustrations of the ovary, or the unitegmic ovules (shouldn’t they be bitegmic in a eudicot angiosperm?), the micropyle (a tiny hole in the integument that the pollen tube grows through), or the papillate inner surface of the ovary. Taken together, their interpretation of all the different organs are consistent, but the uncertainty associated with each part of the fossil makes the final interpretation unreliable in my opinion. It is likely some plant part, but it’s not enough to throw out the last forty years of early angiosperm paleobotany and the last thirty years of molecular phylogenetics, as the headlines may suggest. On the other hand I think Lui and Wang expected that reaction from most of us and they will continue to report on these strange and interesting compression fossils from Northeast China.